v0.9.1

What's Changed

• 102 bug: remove irrelevant params fails for int vaules by @c-w-feldmann in #103
• SmilesToMol: add remove_hydrogens flag by @JochenSiegWork in #108
• 109 chemprop: predictions fail for a single molecule by @c-w-feldmann in #110
• mol2any: get feature names from elements by @JochenSiegWork in #106

Full Changelog: v0.9.0...v0.9.1

v0.9.0

What's Changed

• mol2any: add mol2bool element by @JochenSiegWork in #59
• github: add package source dir to coverage by @JochenSiegWork in #65
• Init refactoring by @frederik-sandfort1 in #70
• 63 include custom filter in experimental by @frederik-sandfort1 in #64
• 72 make chemprop multiclass classification model by @JenniferHem in #73
• 68 makescaffoldgeneric set all carbons to wild card symbols to allow a substrcutre search by @c-w-feldmann in #69
• Error handling sanitize bug by @frederik-sandfort1 in #85
• Inchitomol by @frederik-sandfort1 in #86
• Add --check-only flag to isort by @c-w-feldmann in #88
• docu: update readme, add image, add feature calc notebook by @JochenSiegWork in #90
• add pylint.rc by @c-w-feldmann in #92
• mol2morgan_fingerprint: remove deprecated fp function by @JochenSiegWork in #83
• 81 postpredictionwrapper handle set params without wrapped estimator by @c-w-feldmann in #82
• mol2morgan_fingerprint: fix bug of mismatched indices after folding by @JochenSiegWork in #93
• Add functionality to compare pipelines by @c-w-feldmann in #95
• readme: add link to neural fp paper + make icon smaller by @JochenSiegWork in #97
• Filter update by @frederik-sandfort1 in #67

New Contributors

• @frederik-sandfort1 made their first contribution in #70

Full Changelog: v0.8.6...v0.9.0

v0.8.6

What's Changed

Documentation

• More example notebooks by @JochenSiegWork in #35 and #50

Maintenance

• Switch to new Morgan FP method in tests (part 2) by @c-w-feldmann in #53
• Update requirements_chemprop.txt to allow for chemprop2.0.4 by @c-w-feldmann in #56
• Make code compatible with numpy 1 and 2 by @c-w-feldmann in #57

Bugs

• Fix error caused by using Memory in Pipeline by @c-w-feldmann in #52

Full Changelog: v0.8.5...v0.8.6

v0.8.5

What's Changed

• Update tests to use FingerprintGenerator by @c-w-feldmann in #46
• remove pandas version fix by @c-w-feldmann in #48
• Block logs for physchem descriptors by @c-w-feldmann in #49

Full Changelog: v0.8.4...v0.8.5

v0.8.4

What's Changed

• 27 extend to more fingerprints by @c-w-feldmann in #29
• 31 add kwargs to chempropneuralfp init by @c-w-feldmann in #32
• fix rdkit and chemprop to last functioning versions by @c-w-feldmann in #39
• Fix: moltochemprop not compatible with chemprop 2.0.3 by @c-w-feldmann in #37
• Enhance github actions trigger by @c-w-feldmann in #30
• .github: add test coverage with threshold by @JochenSiegWork in #40
• Remove upper bound for RDKit by @c-w-feldmann in #42
• requirements_chemprop.txt: fix stable chemprop version by @JochenSiegWork in #43
• utils/mol2concatinated_vector: changes for xai by @JochenSiegWork in #33

Full Changelog: v0.8.3...v0.8.4

v0.8.3

Fixes a bug where the sign for loss functions was not used correctly when using the ignored_value_scorer function.

v0.8.2

This release is compatible with sklearn 1.5

v0.8.1

Full Changelog: v0.8.0...v0.8.1

This release also marks the first compatible version of MolPipeline for the https://github.com/basf/neural-fingerprint-uncertainty project.

Release 0.8.0

First public release of MolPipeline.